Having had my first Hello World somewhere back in 2007, for the longest time I have been that old dog that never learns new tricks. One of my regrets now is not adopting any existing PHP frameworks back in late 00s, and instead trying to roll my own. I eventually gave up though, and little by little I have embraced CakePHP by mid 10s.
Still, even with CakePHP, over the last year or so, I have been realising I’m stuck in that 2007 LAMP/LNMP mindset, where you have one website that does everything, without realising that the world had apparently moved on to microservices. I mean of course we send out e-mails using AWS SES, but that’s not really a microservice.
One encounter that has triggered the thoughts resulting in this post is A tale of a dying monolith, where the whole point is getting away from CakePHP. The architecture he ends up with almost fully depends on proprietary technology from 3rd companies, from Prismic to Buddy.works.
Prismic brags about raising 20M to build some kind of slice-based website technology (it’s so many layers of abstraction by now, it seems to have its own learning curve). Good for them! But are they profitable, or will they burn through the money and shut down?
So here I am: on one hand, for the longest time I have been seeing the closed-source, proprietary tech be frown upon. On the other hand, for an MVP-driven development, I’m supposed to drag and drop my no-code AWS SQS module instead of setting up the (seemingly impressive)
I’m leaning towards owning your code and being able to install it on a bare Linux box. Maybe due to the same reasons why I didn’t “trust” a framework 10 years ago. But is it even reasonable now?
Are microservices the future or a buzz word? How does CakePHP fit into that future?
Please share what do you think or feel